簡易檢索 / 詳目顯示

研究生: 鄭宇翔
Cheng, Yu-Hsiang
論文名稱: 考量通用多元件間距限制下的混合元件高度細部擺置
Mixed-Cell-Height Detailed Placement Considering Generic Multi-Cell Spacing Constraints
指導教授: 王廷基
Wang, Ting-Chi
口試委員: 麥偉基
Mak, Wai-Kei
陳勝雄
Chen, Sheng-Hsiung
徐孟楷
Hsu, Meng-Kai
學位類別: 碩士
Master
系所名稱:
論文出版年: 2018
畢業學年度: 106
語文別: 英文
論文頁數: 58
中文關鍵詞: 實體設計細部擺置通用多元件間距限制多倍列高元件混合元件高度設計量產可行性設計接腳可達性
外文關鍵詞: Physical Design, Detailed Placement, Generic Multi-Cell Spacing Constraint, Multi-Row Height Cell, Mixed-Cell-Height Design, Design for Manufacturability, Pin Accessibility
相關次數: 點閱:3下載:0
分享至:
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報
  • 由於製程技術急速成長與晶片製造上的問題,多元件間距限制應運而生。比如,我們可以利用多元件間距限制來處理次10奈米節點下的接腳可達性問題。本論文研究考量了通用的多元件間距限制下的混合元件細部擺置問題。一個直覺上的二元件方法是將每個多元件間距限制切割成多個二元件間距限制,但這樣會造成總元件移動量遠超過需求。因此,欲處理此問題,我們提出一個實際的多元件方法。首先,我們提供了3種方法以分析布局與限制以藉此判斷哪些元件對是最容易拆開。再者,我們應用了一個基於單列動態規劃的方法於各列上,此方法稱作列內移動(IRM),以解決絕大部分的限制違反並最小化總元件移動量或是總線長增加量。有了元件虛擬化(cell virtualization)技術與可移動區域計算(movable region computation)技術,IRM可以被延伸以處理混合元件高度的設計,且無須變更原本動態規劃的方法,只需在成本預估上做一點微小的變更即可達成。最後,我們提供了兩種方法來做全局移動(GM),並保持總元件移動量為最小。眾多的實驗結果顯示與支持我們的多元件方法相當有效率與有效果。


    Multi-cell spacing constraints arise due to aggressive technology scaling and manufacturing issues. For example, we can incorporate multi-cell spacing constraints to tackle pin accessibility problem in sub-10nm nodes. This thesis studies mixed-cell-height detailed placement considering generic multi-cell spacing constraints. A naive 2-cell method is to model each multi-cell spacing constraint as a set of 2-cell spacing constraints, but the resulting total cell displacement would be much larger than necessary. Thus, we aim to tackle this problem and propose a practical multi-cell method. First, we present 3 ways to analyze the initial layout to determine which cell pair in each multi-cell spacing constraint is the easiest to break apart. Secondly, we apply a single-row dynamic programming (SRDP)-based method one row at a time, called Intra-Row Move (IRM), to resolve a majority of violations while minimizing the total cell displacement or wirelength increase. With cell virtualization and movable region computation techniques, our IRM can be easily extended to handle mixed-cell-height designs with only a slight modification of the cost computation in the SRDP method. Finally, we present 2 ways to perform Global Move (GM) to resolve the remaining violations while minimizing the total cell displacement. Versatile experimental results prove and support the efficacy and effectiveness of our approach.

    誌謝 v Acknowledgements vii 摘要 ix Abstract xi 1 Introduction 1 1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Comparison of our Approach and [1]’s . . . . . . . . . . . . . . . . . . 5 1.3 Comparison of our Approach and [2]’s . . . . . . . . . . . . . . . . . . 5 1.4 Our Contributions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 1.5 Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2 Preliminaries 9 2.1 Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.2 Problem Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 3 Our Approach 13 3.1 Overall Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 3.2 Fast Violation Recognition (FVR) . . . . . . . . . . . . . . . . . . . . 14 3.3 Constraint and Layout Analysis . . . . . . . . . . . . . . . . . . . . . . 16 3.3.1 Local Critical Condition (LCC) . . . . . . . . . . . . . . . . . 16 3.3.2 Global Critical Condition (GCC) . . . . . . . . . . . . . . . . . 18 3.3.3 Row-based Local Critical Condition (RLCC) . . . . . . . . . . 22 3.4 Intra-Row Move (IRM) . . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.4.1 The SRDP Model . . . . . . . . . . . . . . . . . . . . . . . . . 23 3.4.1.1 Time Complexity and Space Complexity . . . . . . . 25 3.4.2 Detailed Cost Computations . . . . . . . . . . . . . . . . . . . 27 3.4.2.1 Lazy Cost Method . . . . . . . . . . . . . . . . . . . 29 3.4.3 Detailed Configurations . . . . . . . . . . . . . . . . . . . . . 31 3.4.3.1 The Selection of Rows . . . . . . . . . . . . . . . . . 31 3.4.3.2 The Priority of Rows . . . . . . . . . . . . . . . . . 31 3.4.3.3 Multi-Round IRM . . . . . . . . . . . . . . . . . . . 31 3.4.4 Extensions for Mixed-Cell-Height Designs . . . . . . . . . . . 32 3.4.4.1 Cell Virtualization . . . . . . . . . . . . . . . . . . . 32 3.4.4.2 Movable Region Computation . . . . . . . . . . . . 33 3.4.5 Acceleration . . . . . . . . . . . . . . . . . . . . . . . . . . . 35 3.4.5.1 Parallelization . . . . . . . . . . . . . . . . . . . . . 35 3.4.5.2 Implementation . . . . . . . . . . . . . . . . . . . . 36 3.5 Global Move (GM) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.5.1 Greedy GM (GGM) . . . . . . . . . . . . . . . . . . . . . . . 38 3.5.1.1 Candidate Cells Finding . . . . . . . . . . . . . . . . 38 3.5.1.2 Candidate Empty Spaces Finding . . . . . . . . . . . 38 3.5.1.3 Best Location Determination and Cell Moving . . . . 39 3.5.1.4 Multi-Round GGM . . . . . . . . . . . . . . . . . . 39 3.5.2 ILP-based GM (IGM) . . . . . . . . . . . . . . . . . . . . . . 40 3.5.2.1 Candidate Cells Finding . . . . . . . . . . . . . . . . 40 3.5.2.2 Candidate Empty Spaces and Best Location Finding . 40 3.5.2.3 Best Candidate Empty Spaces Choosing and Cell Moving . . . . . . . . . . . . . . . . . . . . . . . . . . . 40 3.5.2.4 Multi-Round IGM . . . . . . . . . . . . . . . . . . . 41 3.5.2.5 ILP Complexity . . . . . . . . . . . . . . . . . . . . 42 4 Experimental Results 43 4.1 Experiment Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 4.2 Comparison of LCC, RLCC, nLCC, and nRLCC . . . . . . . . . . . . 45 4.3 Comparison of LCC and GCC . . . . . . . . . . . . . . . . . . . . . . 47 4.4 Comparison of Different Number of Conditions . . . . . . . . . . . . . 48 4.5 Comparison of Different Number of Threads . . . . . . . . . . . . . . . 49 4.6 Comparison of GGM and IGM . . . . . . . . . . . . . . . . . . . . . . 50 4.7 2-Cell vs. Multi-Cell Results . . . . . . . . . . . . . . . . . . . . . . . 51 5 Conclusion and Future Works 57 5.1 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 5.2 Future Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 References 59

    [1] D.-W. Huang, “Single-Cell-Height Detailed Placement Considering Multi-Cell
    Spacing Constraints,” Master’s thesis, National Tsing Hua University, Hsinchu,
    Taiwan, 2018.
    [2] C. Han, K. Han, A. B. Kahng, H. Lee, L. Wang, and B. Xu, “Optimal multi-row
    detailed placement for yield and model-hardware correlation improvements in sub-
    10nm vlsi,” in Proceedings of ICCAD, pp. 667–674, 2017.
    [3] B. WINEFELD and S. JILLA, “Understanding physical design constraints in the
    10nm era,” white paper, MENTOR GRAPHICS.
    [4] J. Wang, A. K. Wong, and E. Y. Lam, “Standard cell layout with regular contact
    placement,” IEEE Transactions on Semiconductor Manufacturing, vol. 17,
    pp. 375–383, Aug. 2004.
    [5] S. Dobre, A. B. Kahng, and J. Li, “Mixed cell-height implementation for improved
    design quality in advanced nodes,” in Proceedings of ICCAD, pp. 854–860, 2015.
    [6] S.-H. Baek, H.-Y. Kim, Y.-K. Lee, D.-Y. Jin, S. Park, and J.-D. Cho, “Ultra-high
    density standard cell library using multi-height cell structure,” in Proceedings of
    SPIE, pp. 72680C–1–72680C–8, 2008.
    [7] Y. Du and M. D. F. Wong, “Optimization of standard cell based detailed placement
    for 16 nm finfet process,” in Proceedings of DATE, 2014.
    [8] Y. Ding, C. Chu, and W.-K. Mak, “Pin accessibility-driven detailed placement
    refinement,” in Proceedings of ISPD, pp. 133–140, 2017.
    [9] Y.-Y. Wu and Y.-W. Chang, “Mixed-cell-height detailed placement considering
    complex minimum-implant-area constraints,” in Proceedings of ICCAD, pp. 65–
    72, 2017.
    [10] V. Yutsis, I. S. Bustany, D. Chinnery, J. R. Shinnerl, and W.-H. Liu, “ISPD 2014
    benchmarks with sub-45nm technology rules for detailed-routing-driven placement,”
    in Proceedings of ISPD, pp. 161–168, 2014.
    [11] I. S. Bustany, D. Chinnery, J. R. Shinnerl, and V. Yutsi, “ISPD 2015 benchmarks
    with fence regions and routing blockages for detailed-routing-driven placement,”
    in Proceedings of ISPD, pp. 157–164, 2015.
    [12] N. K. Darav, I. S. Bustany, A. Kennings, and R. Mamidi, “ICCAD-2017 cad contest
    in multi-deck standard cell legalization and benchmarks,” in Proceedings of
    ICCAD, pp. 867–871, 2017.
    [13] “Si2 OpenAccess.” http://projects.si2.org/oac_index.php. Accessed: 2018-
    04-01.
    [14] A. B. Kahng, P. Tucked, and A. Zelikovsky, “Optimization of linear placements
    for wirelength minimization with free sites,” in Proceedings of ASPDAC, pp. 241–
    244, 1999.
    [15] “Gurobi Optimization.” http://www.gurobi.com/. Accessed: 2018-04-01.
    [16] “OpenMP.” http://www.openmp.org/. Accessed: 2018-04-01.
    [17] “OpenCores.” https://opencores.org/. Accessed: 2018-04-01.
    [18] “Synopsys Design Compiler Graphical.” https://www.synopsys.com/
    implementation-and-signoff/rtl-synthesis-test/design-compiler-graphical.
    html. Accessed: 2018-04-01.
    [19] “NanGate FreePDK15 Open Cell Library.” http://www.nangate.com/?page_
    id=2328. Accessed: 2018-04-01.
    [20] “Cadence Encounter Digital Implementation.” https://www.cadence.com/
    content/cadence-www/global/en_US/home/training/all-courses/82160.
    html. Accessed: 2018-04-01.
    [21] “ICCAD 2017 Contest Problem C - Multi-Deck Standard Cell Legalization.”
    http://cad-contest-2017.el.cycu.edu.tw/Problem_C/default.html. Accessed:
    2018-04-01.
    [22] “OpenGL.” https://www.opengl.org/. Accessed: 2018-04-01.

    QR CODE